Has anyone considered whether the timer resolution of 1 usec is adequate for oneshot 125?
truthfully i donāt know alot about this this software yet, but i have had alot of issues with this build , and i just happen to stumble apon this info by trial and error
updating findings oneshot125 setup for qgc is supposed to be pwm_max 250, pwm_min 125, pwm_disarm 100 , pwm_rate 4000hz , all mosted be forced saved , calibation works as well
info found 13 minutes into video
Some progress on that side, or one does not give up easily.
I was able to calibrate ESCs with Oneshot125, but ā¦
I tried again and it did not work, obviously it wasnāt before it shouldnāt now. Unless I did a mistake.
I tried setting PWM_RATE at 400 instead of 4000 and this time I did the calibration moving throttle stick from high when powering the quad with the battery, to low after the first sound. It did the calibration and there was an audio feedback.
Now everything works except that it will not arm in Position Hold. I was almost sure it did before, am I wrong?
Motors seems to be more responsive, but difficult to say without propeller.
Is it really working in Oneshot125?
It seems DampenLight is not enabled. Test was to move right stick to a position where one motor does not spin. There was no braking effect moving that motor by hand. I guess then with a propeller it will continue to spin. Makes sense?
Any input appreciated!
Ric
i have been through all those settings man it pwm_rate is 4000hz i promise you unless something change with the build between yours and mine
have u pluged in your ESCs to a flash tool
to see if lightdamped is turned in ,also if u do calibration in 400hz it will sound off and u have to move the throttle to get it to respond but this just tricking u into thinking its working , but its not and when u do this it puts your ESCs into normal mode thats why lightdamped gets turned off
Thanks for your inputs.
I tried to program the ESCs with a programming card.
If I set control frequency to 22KHz then the ESCs will calibrate both with PWM set 400 and 4000.
But with the last one they will not arm.
A problem I can see with that setting is ESCs becoming warm.
I programmed brake on, always with the programming card and now I feel the braking effect.
I guess that if min and max PWM are set at 125 and 250 it should work as Oneshot125 (syncronized pulse), just refreshed only at 400Hz and not at 4000Hz. Am I completely out of track? 
The motors response seems good anyway.
Now itās slope time ā¦
active brake setting is independent from oneShot, no matter 400 or 4000. What ESC you have and which BL/S firmware version? seems you have an elder one as active brake is ON by default in newer releases.
4000 should work. Probably should try some other ESCs, or later BL firmware.
ESCs are Emax Oneshot125 20A. I canāt access the firmware so donāt know the version. Seems those ESCs have a strange version of One shot.
thatās more old model. It works fine but have known issues of FET management in firmware. It generates heat even during motor idle and drain your battery, and quite easy to get on fire. I burned 3 of them(real flames!) even with no props on. So be careful.
May worth trying some newer version
Thanks, so you confirm what Iāve read around: those are crappy ESCs.
Will eventually swap them, as now they donāt work correct anymore. I seem was able to corrupt something.
when u tried to read the firmware with the tool what chip set did u try?
maybe there atmel and not silabs
@chiloschista Iād appreciate if if you could try https://github.com/PX4/Firmware/pull/6616
It implements oneshot correctly and is flight tested on a Pixracer. Note that the only parameter change required is to set PWM_RATE to zero. Leave PWM min/max/disarmed at the ānormalā values.
Please let me know if you have any questions about setup after reading the PRs.
āMark
I would really like to do it, but that multirotor does no more pass any ground test. Seems I was able to corrupt something (ESCs firmware?).
Iām planning to swap them. Any suggestion about thoose?
Flyduino KISS 24A? they tell every protocol will automatically recognized, from standard PWM to Oneshot 125 and 42.
DYS XSD20A?
DYS XS20A?
By the way where do I find #6616? Still not very used to Github and please tell me I donāt have to use Sourcetree, itās a pain 
As soon as the multirotor passes again ground tests I will go for it.
Ric
I checked the ESCs with a servo tester and just one is not working properly. It only get the motor to partial revs.
That one was the only one with BEC connected to Pixracer. I left the voltage pin for servo rail feed, in the case I need it, but now I noticed it will power the FC also. Should it not be separed from main connector as the main voltage does not feed the servos rail? Is that a problem?
try the ,fvt littlebee 20a they seem to work good
I have only tested 2 types of ESC with oneshot125,
http://www.buddyrc.com/spedix-30a-high-voltage-esc-with-blheli-s-program.html with BlHeli firmware (new)
and
http://www.buddyrc.com/ztw-spider-20a-opto-multirotor-esc.html with SimonK (1 year old)
Regarding testing of a PR, the easiest way is probably to just clone my repo, checkout the add_oneshot_mode branch and build it locally. Let me know if that doesnāt make sense, and I can provide more detailed instructions.
just to repost my esc info ,i flashed afro 30a opto ESCs with blheli 14.9 , and they were simonk firmware when bought comunication frequency of 1khz